Kitz ADSL Broadband Information
adsl spacer  
Support this site
Home Broadband ISPs Tech Routers Wiki Forum
 
     
   Compare ISP   Rate your ISP
   Glossary   Glossary
 
Please login or register.

Login with username, password and session length
Advanced search  

News:

Pages: [1] 2 3 ... 9

Author Topic: Cable Pairs - A Question  (Read 35390 times)

snadge

  • Kitizen
  • ****
  • Posts: 1299
Cable Pairs - A Question
« on: April 20, 2012, 11:12:21 PM »

as some of you know it took 4 engineer visits to locate a battery contact fault in my line.

my house has a Junction Box on it on the wall (JB1) - it has 4 pairs going along to another Junction Box on end house 6 doors down (JB2), from there the cables travel 20m to a DP in a duct (DUCT).
so route is NTE-5 -> JB1 -> JB2 -> DUCT (then onto cab then exchange)

now engineer no.2 changed me onto a clean pair in same cable at JB2 (6 doors down) but could not access the duct to hook me up in there.
engineer no.3 goes into duct and hooks that pair up from JB2.
engineer no.4 finds the fault between JB1 and JB2 - he switches me onto a spare pair in that run...

now my pairs from JB1 to JB2 to DUCT have all been changed, the engineers did not leave site once and my service was working when they connected up and tested... now does this mean from the DUCT down to exchange my pair is the same as before? - im being told that he must of changed D-side and E-side tie pairs so im now on a new line right down to the exchange (even though its in same cable) - but Iam not convinced, I think its only that short 70m run thats been changed, the engineers did not visit any cabs or the exchange, am I right in thinking this? or has my line been changed over entirely?

that 70m-90m run (1100m line length or thereabouts) shouldnt make much difference to my connection? our lass says the phone is the clearest its ever been...

Ive been told that its likely there is still a fault on the line (probably at exchange because it all happened when sky switched us over) but convincing sky to do anything about it will be impossible, my 18Mb connection is now 11Mb

anyway - am i right in saying that about only the 70-90m run has been changed?



thanks in advance
« Last Edit: April 21, 2012, 04:02:01 PM by snadge »
Logged
Zyxel VMG8924-B10A <-> PlusNET <-> Eci DSLAM
Sync: 64,000k <-> INP: ON <-> FTTP available :) 900Mbps +5G EE Mobile

waltergmw

  • Content Team
  • Kitizen
  • *
  • Posts: 2774
Re: Cable Pairs - A Question
« Reply #1 on: April 21, 2012, 12:11:14 AM »

@ Snadge,

BS Might add some more here, but here's a start.

If a lengthy operation was accomplished you would expect to see the engineer plug a tone generator into your master socket and use that to find, and then change your pair. If the engineer's van did not leave your site and then return later then only local changes have been made.

Kind regards,
Walter

Logged

snadge

  • Kitizen
  • ****
  • Posts: 1299
Re: Cable Pairs - A Question
« Reply #2 on: April 21, 2012, 12:17:05 AM »

thanks walter

yeah he used a TG and JDSU - none of them left the site, i watched them like a hawk hehe
Logged
Zyxel VMG8924-B10A <-> PlusNET <-> Eci DSLAM
Sync: 64,000k <-> INP: ON <-> FTTP available :) 900Mbps +5G EE Mobile

asbokid

  • Kitizen
  • ****
  • Posts: 1286
    • Hacking the 2Wire
Re: Cable Pairs - A Question
« Reply #3 on: April 21, 2012, 12:18:04 AM »

Hello snadge,

BlackSheep is really the expert to comment on the routing, cabling, tie-pairs, etc..

Your graphs interest me though..



The graphing quirks of DMT make it difficult once again to directly compare.  The first graph from your former o2/BE service, appears to show an ADSL2+ G992.5 Annex M service, whereas the Sky service is plain old Annex A. [1]    Yet we would expect the downstream performance to be favourable in Annex A since more bits are allocated to the downstream than in Annex M.

The low definition of the attenuation graphs make direct comparisons difficult. However the two attenuation plots for o2 and for Sky, look very similar.    At the highest frequencies in the ADSL2+ spectrum, both graphs indicate an attenuation of just over -60dB.

What the graphs do clearly illustrate is that the average bit depth per tone is significantly less with Sky than it was with o2/BE.

The notches in the bit-loading and in the SNR graphs that were present in the o2 service have persisted, and remain focused on the same spectral regions, albeit in exaggerated form. Those notches are much more pronounced with the Sky service and are resulting in greater losses in bandwidth.  As an aside, the notches look like they are caused by RFI-related noise, which could perhaps be caused by equipment in your household.

For whatever reason, the transceiver in the Sky linecard is not so competent as the o2 linecard was at modulating signals that your modem can recover.

cheers, a

[1] http://en.wikipedia.org/wiki/ITU_G.992.5_Annex_M
[2] http://www.skyuser.co.uk/forum/sky-broadband-exchange-information/47077-exchange-supports-asdl2-annex-m.html
Logged

snadge

  • Kitizen
  • ****
  • Posts: 1299
Re: Cable Pairs - A Question
« Reply #4 on: April 21, 2012, 12:39:44 AM »

i have higher def ones for you below

many thanks for the insight asbokid i really appreciate it, i noticed those spikes are in both graphs but the whole spectrum on sky is weakened, not just in those ranges where the RFI spikes are

I didnt think I was on Annex M mind? I never had 3Mb upload, my stats where 1,300k up and 18,600k down with 3db SNRM (16,800k was typical sync rate for SNRM of 6.5db) - but your right cos on SKY i could achieve 18,013k with 7db SRNM - so i was getting more speed than BE/o2 in that respect.

I have been on sky 6 weeks prior to this and acheived 18Mb on their MSAN , when my phone followed it across all this crap happened and i dont believe it was a coincidence, within hours of switch we had problems with noise, dropouts, phone ringing, voices on line, now this is supposedly fixed cos of a fault in cable outside my house?

there is an alarm box we have that makes a humming sound not far from router...that may be source of those 'spikes' , Tomorrow I will see if I can disable it, remove battery and turn off power to boiler, cooker and everything except ground floor sockets and then ensure nothing is plugged in but the router and laptop and re-check

but what can be causing the overall degradation of that signal over the entire spectrum? like I say i had 18Mb sync (7db NM) for 6 weeks prior and it was fine until hours after the switch, dunno if its related lol,,, my next door neighbour has a wireless door bell, it rings everytime a car drives past his house??? lol.... just thought I would mention it haha

here is the SKY one currently: upstream LA seems to be fluctuating between 10-16db and I think i have high errors , especially on upstream




and here is the o2/BT one:



thanks
« Last Edit: April 21, 2012, 12:50:59 AM by snadge »
Logged
Zyxel VMG8924-B10A <-> PlusNET <-> Eci DSLAM
Sync: 64,000k <-> INP: ON <-> FTTP available :) 900Mbps +5G EE Mobile

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 32304
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Cable Pairs - A Question
« Reply #5 on: April 21, 2012, 01:17:43 AM »

A quick comment before Black Sheep arrives.

What you are discussing here is nothing to do with "Tie-Pairs". The are Something Completely Different . . . (b*cat plays The Liberty Bell March, by J P Sousa, on the electronic organ.  ::)  )

From the description of the work carried out, you have had your service swapped to a different pair between the first junction box (JB1 in your description) and the DP.

The rest of your D-side pair and your E-side pair appear not to have been swapped.
« Last Edit: April 21, 2012, 02:08:51 AM by burakkucat »
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

asbokid

  • Kitizen
  • ****
  • Posts: 1286
    • Hacking the 2Wire
Re: Cable Pairs - A Question
« Reply #6 on: April 21, 2012, 01:28:48 AM »

Hi again, Snadge,

i noticed those spikes are in both graphs but the whole spectrum on sky is weakened, not just in those ranges where the RFI spikes are

One thing that struck me... except for the change of service provider, are you using the same modem as you did with the o2/Be service?

Quote
I didnt think I was on Annex M, mind? I never had 3Mb upload, my stats were 1,300k up and 18,600k down with 3db SNRM (16,800k was typical sync rate for SNRM of 6.5db) -

Ahh. that pretty much confirms that you were on Annex M.   For Annex A, the maximum attainable upstream rate (sync rate) is 1024kbps, which you exceeded with 1300kbps, which indicates that you were getting an Annex M service with o2/Be.

It's quite possible that DMT was wrongly reporting Annex M but it is apparently an option with o2's 'Pro' package, and Annex M has reportedly been provided as a gratis service to those on o2's cheaper packages.

Quote
but you're right cos on SKY i could achieve 18,013k with 7db SNRM - so i was getting more [downstream] speed than BE/o2 in that respect.

Ahh.. that pretty much seals it then.

Quote
I have been on sky 6 weeks prior to this and achieved 18Mb on their MSAN , when my phone followed it across all this crap happened and i don't believe it was a coincidence, within hours of switch we had problems with noise, dropouts, phone ringing, voices on line, now this is supposedly fixed cos of a fault in cable outside my house?

That definitely sounded like the engineer introduced a split pair, although the performance remained surprisingly high in spite of it.

Quote
there is an alarm box we have that makes a humming sound not far from router...that may be source of those 'spikes' , Tomorrow I will see if I can disable it, remove battery and turn off power to boiler, cooker and everything except ground floor sockets and then ensure nothing is plugged in but the router and laptop and re-check

Yeah, do try that. Very good idea.

Quote
but what can be causing the overall degradation of that signal over the entire spectrum? like I say i had 18Mb sync (7db NM) for 6 weeks prior and it was fine until hours after the switch, dunno if it;s related lol,,, my next door neighbour has a wireless door bell, it rings every time a car drives past his house??? lol.... just thought I would mention it haha

Wretched neighbours! 

If you can get the Quiet Line Noise (QLN) data from the modem (what model is it?)  that would allow a lot more detective work on the cause of the overall degradation of your signal, and the possible cause(s) of those notches.


Thanks for posting!  I know it's never nice to have a fault but as perverted as it might see, some people enjoy studying others' problems!  My GP is one of them.  He seems delighted to discover something 'wrong'!   "How fascinating! Your leukocyte count is astonishing, the highest I've ever seen! My goodness! You've made my day! Is there a history of early death in your family? "

cheers, a
« Last Edit: April 21, 2012, 02:09:32 AM by asbokid »
Logged

snadge

  • Kitizen
  • ****
  • Posts: 1299
Re: Cable Pairs - A Question
« Reply #7 on: April 21, 2012, 02:11:58 AM »

@ B'kat - yes.. just "pairs" then is it? :) ...thanks for correcting me, i'll try and change the title and wording :)

@ asbokid -

Quote
One thing that struck me... except for the change of service provider, are you using the same modem as you did with the o2/Be service?

yup same one.. Netgear GT


Quote
Ahh. that pretty much confirms that you were on Annex M

wierd that... who knew? lol ... mind on SKY (annex a) Iam acheiving 1150-1160k ..?


Quote
That definitely sounded like the engineer introduced a split pair, although the performance remained surprisingly high in spite of it.

thats what someone else who knows his stuff said, when it happened the broadband was anywhere from 7Mb-13Mb , when an engineer went to fix it the phone went dead completely and speed stuck at 3Mb - stayed like that for 4 or 5 days cos sky gave me the run-a-round, then 2nd engineer couldnt do anything, 3rd engineer swapped the pair in duct but problems were still there (how didnt he pick up this fault?) and speeds were same as now, then 4th engineer swaps pair from my house to end house (so now the run from JB1 to JB2 to DUCT is all on a new pair in same cable) and the problematic noisey phone and SNR spikes causing dropouts stop - but broadband remains slow with upstream line attenuation fluctuations 10-16db and i think Iam getting a lot of upstream errors but unsure...


Quote
If you can get the Quiet Line Noise (QLN) data from the modem (what model is it?)  that would allow a lot more detective work on the cause of the overall degradation of your signal, and the possible cause(s) of those notches.

its a Netgear DG834GT (DGteam 1018) - I also have the Sagemcom 2504N (which is a better router but is locked out) and I have a Speedtouch 585v6 , this sounds promising? I never heard of QLN values in router...love it when i learn something new :)  -can my GT access this value via telnet? if so how do I do that? cheers


Quote
Thanks for posting!  I know it's never nice to have a fault but as perverted as it might see, some people enjoy studying others' problems!  My GP is one of them.  He seems delighted to discover something 'wrong'!   "How fascinating! Your leukocyte count is astonishing, the highest I've ever seen! Is there a history of early death in your family?"

hahaha - that cheered me up, hey the pleasure is all mine... you have been a massive help as you have pointed out a lot of stuff I wasnt aware of and looking forward to my tests tomorrow and hopefully getting this QLT value - Sky's Zoe Quiet Line Test said "no echo and a little noise" if thats any help - my A/C Balance on line was 63db (if thats any help lol) - the engineers JDSU reported the pair quality test was good...but that would be the physical aspect of it I suppose.

thanks again for your help, I look forward to posting my results tomorrow - I hope it is in my home because then I know I can rectify it...

CHEERS :)
Logged
Zyxel VMG8924-B10A <-> PlusNET <-> Eci DSLAM
Sync: 64,000k <-> INP: ON <-> FTTP available :) 900Mbps +5G EE Mobile

asbokid

  • Kitizen
  • ****
  • Posts: 1286
    • Hacking the 2Wire
Re: Cable Pairs - A Question
« Reply #8 on: April 21, 2012, 03:00:42 AM »

Quote
If you can get the Quiet Line Noise (QLN) data from the modem (what model is it?)  that would allow a lot more detective work on the cause of the overall degradation of your signal, and the possible cause(s) of those notches.

its a Netgear DG834GT (DGteam 1018) - I also have the Sagemcom 2504N (which is a better router but is locked out) and I have a Speedtouch 585v6 , this sounds promising? I never heard of QLN values in router...love it when i learn something new :)  -can my GT access this value via telnet? if so how do I do that? cheers

The modem that DMT queried for those line stats - that is the modem with the Broadcom BCM6348 chipset - must have a telnet interface in the firmware. There must also be a tool called adslcfg a.k.a. adslctl or even just adsl that is obtaining the tonemap data from the DSP driver for SNR, QLN, HLOG and Bit Loading.

The tool, whatever it is called in your firmware, will produce output like this..

Code: [Select]
# xdslcmd info --SNR
...
Tone number      SNR
   0 0.0000
   1 0.0000
   2 0.0000
   3 0.0000
...
   31 0.0000
   32 0.0000
   33 43.5000
   34 44.7500
...
   508 29.6875
   509 29.5625
   510 30.1875
   511 29.0000
#

DMT and RouterStats graphically plot that data.  There are four main datasets available from Broadcom chipset modems:

Quiet Line Noise (dBm/Hz) or QLN - the DSL equivalent to the Quiet Line voice test.
Signal Noise Ratio (dB) or SNR - recorded on a per subcarrier basis
Attenuation(dB) or HLOG/HLIN - recorded on a per subcarrier basis (HLOG is a logarithmic (decibel) value for attenuation. HLIN is linear scale)
Bit Loading - number of bits loaded per subcarrier

Quote
I look forward to posting my results tomorrow - I hope it is in my home because then I know I can rectify it...

Indeedy.

cheers, a

« Last Edit: April 21, 2012, 03:07:24 AM by asbokid »
Logged

snadge

  • Kitizen
  • ****
  • Posts: 1299
Re: Cable Pairs - A Question
« Reply #9 on: April 21, 2012, 04:01:05 PM »

well I disabled power to house alarm, boiler, cooker, turned on only ground floor sockets and only the router and modem plugged in, everything else around the house unplugged from wall, phone removed from filtered faceplate so its just the router

and results were identical:- 
- checkout the fluctuations of upstream Line Attenuation, thats not normal is it? ive seen it between 10-16 in last 24 hours

WITH EVERYTHING ON AS NORMAL:




WITH EVERYTHING OFF:




I think that confirms its not REIN in my house but is external, as I had those spikes before all my troubles its safe to say that THEY are not the cause of my current speed loss either

I used DMT to send command via Telnet, I dont know the command but playing around with it I managed to find this:
adslctl info --QLT

and the returned info was:

adslctl info --QLT
adslctl: too many options
#


was that the wrong command?  Iam unfamiliar with telnet (as you can see)


I also think Iam getting a lot of upstream errors but not sure, struggling to get someone to tell me if they are high or not and how to tell for myself :(


thanks for the help BTW :)

EDIT: router has dropped twice on its own in last 5 minutes, i may be wrong but iam sure errors are high and upstream Line Attenuation has creeped up again to 15.4db - so today its been 11db , 13db, 14db ,15.4db (not in that order , it seems to be random rather that 'climbing')
« Last Edit: April 21, 2012, 04:21:51 PM by snadge »
Logged
Zyxel VMG8924-B10A <-> PlusNET <-> Eci DSLAM
Sync: 64,000k <-> INP: ON <-> FTTP available :) 900Mbps +5G EE Mobile

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 32304
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Cable Pairs - A Question
« Reply #10 on: April 21, 2012, 06:46:32 PM »

The measured attenuation of the metallic-loop should not vary.  :no:

The fact that it is doing so does not indicate REIN but a physical defect somewhere in the pair between your NTE5/A and (say) the HDF in the exchange.  :-X

Hopefully Black Sheep may be able to add further comments or suggestions?  ???
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

GigabitEthernet

  • Kitizen
  • ****
  • Posts: 2082
Re: Cable Pairs - A Question
« Reply #11 on: April 21, 2012, 08:17:12 PM »

On the DG834GT, you should be able to type "adslctl info --SNR" to get the information asbokid is looking for.
Logged

asbokid

  • Kitizen
  • ****
  • Posts: 1286
    • Hacking the 2Wire
Re: Cable Pairs - A Question
« Reply #12 on: April 21, 2012, 08:34:57 PM »

well I disabled power to house alarm, boiler, cooker, turned on only ground floor sockets and only the router and modem plugged in, everything else around the house unplugged from wall, phone removed from filtered faceplate so its just the router

and results were identical:- 
- checkout the fluctuations of upstream Line Attenuation, thats not normal is it? ive seen it between 10-16 in last 24 hours
...
I think that confirms its not REIN in my house but is external, as I had those spikes before all my troubles its safe to say that THEY are not the cause of my current speed loss either

That's a shame. Maybe the next step is to note the precise frequencies of the noise and with an appropriate radio receiver, try to track down the source.   It would be best first to obtain accurate tonemaps for Quiet Line Noise (QLN), etc..

Quote
I used DMT to send command via Telnet, I dont know the command but playing around with it I managed to find this:
adslctl info --QLT

and the returned info was:

adslctl info --QLT
adslctl: too many options


Put your telnet client into 'log to file' mode..  then issue each of the following commands in turn. The output will run to several thousand lines of raw data.  Once it's graphed it's much easier to study. 

Code: [Select]
adslctl info --vendor
adslctl info --show
adslctl info --SNR
adslctl info --QLN
adslctl info --Hlog
adslctl info --Bits

You're welcome to email or PM it, or zip the output and attach it to a post, maybe?

cheers, a
Logged

Black Sheep

  • Helpful
  • Addicted Kitizen
  • *
  • Posts: 5249
Re: Cable Pairs - A Question
« Reply #13 on: April 22, 2012, 03:14:12 PM »

as some of you know it took 4 engineer visits to locate a battery contact fault in my line.

my house has a Junction Box on it on the wall (JB1) - it has 4 pairs going along to another Junction Box on end house 6 doors down (JB2), from there the cables travel 20m to a DP in a duct (DUCT).
so route is NTE-5 -> JB1 -> JB2 -> DUCT (then onto cab then exchange)

now engineer no.2 changed me onto a clean pair in same cable at JB2 (6 doors down) but could not access the duct to hook me up in there.
engineer no.3 goes into duct and hooks that pair up from JB2.
engineer no.4 finds the fault between JB1 and JB2 - he switches me onto a spare pair in that run...

now my pairs from JB1 to JB2 to DUCT have all been changed, the engineers did not leave site once and my service was working when they connected up and tested... now does this mean from the DUCT down to exchange my pair is the same as before? - im being told that he must of changed D-side and E-side tie pairs so im now on a new line right down to the exchange (even though its in same cable) - but Iam not convinced, I think its only that short 70m run thats been changed, the engineers did not visit any cabs or the exchange, am I right in thinking this? or has my line been changed over entirely?

that 70m-90m run (1100m line length or thereabouts) shouldnt make much difference to my connection? our lass says the phone is the clearest its ever been...

Ive been told that its likely there is still a fault on the line (probably at exchange because it all happened when sky switched us over) but convincing sky to do anything about it will be impossible, my 18Mb connection is now 11Mb

anyway - am i right in saying that about only the 70-90m run has been changed?



thanks in advance

Terminology is key to understanding what is happening. Who "is telling you" that the E and D-sides have been changed ?? Also, for clarification, 'tie-pairs' only exist in Telephone Exchanges. The correct wording is just 'pairs', when referring to the Network cabling. It does make a difference when reading event logs on here. ;) :)

From your description, it sounds like the engineers have only changed the pair within the 90mtr cable runs you mention. However, if that is where the fault was found to be, they would not just continue changing the pair all the way back to the Exchange. It wouldn't be cost-effective, time-effective, or serve any purpose.

If a 'pair' is changed on the D-side network from the PCP to the DP, that is always termed a 'D-side swap/change', if it's from an underground joint to another underground joint, or from an underground joint anywhere along the D-side cable run to the DP, this is termed as 'Pair changed in length'. The former is a 81.7P clear code, the latter an 82.1P clear code. Your ISP should have a history of which code was used. We use various codes as an audit trail so our Quality Auditos can perform spontaneous checks on our workmanship.

HTH bud.
Logged

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 32304
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Cable Pairs - A Question
« Reply #14 on: April 22, 2012, 05:06:16 PM »

<snip>

From your description, it sounds like the engineers have only changed the pair within the 90mtr cable runs you mention. However, if that is where the fault was found to be, they would not just continue changing the pair all the way back to the Exchange. It wouldn't be cost-effective, time-effective, or serve any purpose.

If a 'pair' is changed on the D-side network from the PCP to the DP, that is always termed a 'D-side swap/change', if it's from an underground joint to another underground joint, or from an underground joint anywhere along the D-side cable run to the DP, this is termed as 'Pair changed in length'. The former is a 81.7P clear code, the latter an 82.1P clear code. Your ISP should have a history of which code was used.

<snip>

 :thumbs:  Thanks for your clarification, BS.
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.
Pages: [1] 2 3 ... 9